Pediatric Pulmonology 11 :1-7 (1991)

Original Articles -

Evaluation of Nasal Impedance Using the Forced Oscillation Technique in Infants

K.N. Desager, M D , ~ M. Willernen,’ H.P. Van Bever, M D , ~ W. De Backer, MD, P~D,’ and P.A. Vermeire, M D ~

Summary. By applying oscillations to the respiratory system through a rigid face mask, the infant-adapted Landser forced oscillation technique measures impedance of the total respiratory system including the nose, at frequencies from 4 to 52 Hz. The present study was aimed at evaluating nasal impedance in infants from consecutive forced oscillation measurements through both nostrils and each nostril separately, using a simple electrical model. In 30 asthmatic infants with varying degrees of nasal obstruction, aged 1-16 months, calculated nasal resistance (R,) at 24 Hz ranged from 1 to 16 cm H,O. L-’ s. The ratio of R, to total respiratory system resistance varied between 1 and 48% (mean: 16%). In seven non-asthmatic infants, aged 0-12 months, R, was between 1 and 11 cm H,O . L-’ . s. Nasal patency (evaluated clinically) was correlated with the calculated R, (P i 0.05). R, showed almost no frequency dependence between 24 and 48 Hz as demonstrated by a mean slope of -0.09 f 0.08 cm H,O . s2/L for the asthmatic and of -0.08 ? 0.07 for the non-asthmatic infants. In seven of the asthmatic infants the differences between two R, determinations at a 45 min interval ranged from -1.7 to 3.8 cm H20. L-’ s-‘ at 24 Hz and from -3.6 to 1.0 at 48 Hz. Changes in R, did not correlate with changes in total respiratory system resistance (P > 0.05). In conclusion, nasal impedance can be approximated from three consecutive measurements through both nostrils and through each nostril separately. Pediatr Pulmonol. 1991 ; 11 :1-7.

Key words: Infants with asthma vs. controls; ratio of nasal to total respiratory system resistance; variability of nasal resistance.

INTRODUCTION

Measurements of total pulmonary resistance in infants have been performed with various methods. ’,* Since infants breathe preferably through their noses, these measurements include the resistance of nasal air pas- sages. Infants with asthmatic symptoms frequently suffer from associated rhinitis and therefore nasal resistance (R,) may account for a considerable percentage of measured total pulmonary resistance. Relatively few measurements of R, have been performed in infant^,^-^ probably because of the technical difficulties involved in adapting usual methods for measuring R, in adults, e.g., anterior and posterior rhinomanometry. The purpose of the present investigation was to perform measurements with the pseudo-random noise (PRN) oscillation tech- nique through both nostrils and through each nostril separately and to calculate from these measurements nasal impedance, using a simple electrical model and assumptions outlined by Lacourt and P ~ l g a r . ~ To evalu- ate time related changes in nasal impedance, two sequen- tial determinations were performed in a subgroup of seven infants. 0 1991 Wiley-Liss, Inc.

MATERIALS AND METHODS Measurements were performed using the PRN forced

oscillation technique as described in detail previously. * Briefly, a PRN signal, generated by a loudspeaker and containing all harmonics of 4 Hz up to 52 Hz,9 is applied to the respiratory system. Differential pressure across a Lilly type pneumotachograph and mouth pressure are measured with identical differential pressure transducers (Validyne MP 45-l), with a range of 2 2 crnH,O. Mouth pressure and flow signals are fed into a Fourier analyzer. The impedance of the respiratory system, calculated at 4, 8, . . . 52 Hz, is partitioned into a real part or resistance (R) and an imaginary part or reactance (X). A coherence

From the Departments of Respiratory Medicine’ and Pediatrics,2 University Hospital of Antwerp, Edegem, Belgium.

Received May 28 , 1990; (revision) accepted for publication January 17, 1991.

Address correspondence and reprint requests to Dr. P.A. Vermeire, Department of Respiratory Medicine, University Hospital Antwerp, 8-2650 Edegem, Belgium.

2 Desager et al.

function, evaluating the amount of noise and alinearities present in the measured signal, was calculated for all harmonics from 4 to 52 Hz. Only values with a coherence function greater than 0.94 were accepted.

The technique has been adapted in our laboratory for use in infants.” A 60 cm long extension tube was installed to perform measurements in the supine position with a bias flow of 6 L/min to reduce CO, retention. A well-fitting stiff face mask was used to apply oscillations to the respiratory system of nose-breathing infants.

Impedance values were obtained when the infant was breathing through both nostrils, as well as through the left and right nostril separately. Lips were closed and fixed with adhesive tape. Successive measurements were al- ways performed in the same order. Nostrils were oc- cluded with a conical silicone plug of appropriate size. Care was taken to avoid deviation of the nasal septum. Values of five measurements, each lasting approximately 16 s, were averaged.

A simple electrical model was used to approximate nasal impedance (Fig. 1). Nasal air passages were con- sidered to have an R and an X, representing inertance or compliance of the nasal passage or both; they were arranged in parallel with each other and in series with R,, and X,, representing R and X, respectively, of the total respiratory system without the nose. Using this model the mathematical relationship between measured impedance values and their components can be expressed by three formulas:

1. Measurements through both nostrils yield the im- pedance of the total respiratory system including the nose (Ztot) and are related to the impedance of the left nostril (Z,), of the right nostril (Z,) and of the respiratory system, nose excluded (Z,,).

2 and 3. Measurements through the left and then right nostrils yield Z,,tot, Z,,,,,, i.e., impedance of the respi- ratory system, including the left and then right, nostrils:

For each of the 13 frequencies these equations can be solved to yield R and X value for the nose (R,, X,) and for the rest of the respiratory system (R,,, X,,).

In 32 infants, aged 1-16 months, with recurrent periods of wheezing, rhinitis and cough, and in seven non-asthmatic infants, aged 0-12 months, the three series of measurements were performed in supine posi- tion after sedation with 100 mg/kg chloral hydrate p.0. Nasal patency of both groups of infants was evaluated

nose R X

0 Rrs Xrs .

left nostril occluded - .

-3) . I

I right nostril occluded

Fig. 1. Electrical model representing measurements through one and both nostrils. For details see text.

clinically prior to the lung function testing and was classified as being “normal” or “reduced.” Informed parental consent was obtained and the study was ap- proved by the Hospital Ethical Committee. In 7 of the 32 asthmatic infants two nasal impedance values were determined sequentially within a time interval of 45 min.

Mean R, of the two groups with different nasal patency was evaluated statistically using a Wilcoxon rank sum test. A paired t test was used to evaluate the difference between the frequency dependence of measurements performed through the nostrils and of the nasal resis- tance.

RESULTS In 2 of the 32 infants it was not possible to occlude one

nostril because the other was totally obstructed and therefore no nasal impedance values could be calculated.

In the other 30 infants, R, varied between 0.8 and 16.6 cm H,O L-’ s at 24 Hz (Fig. 2) and between 0.1 and 5.6 at 48 Hz, whereas R of the total respiratory system including the nose (Rtot) ranged from 17 to 56 cm H,0 * L-l s at 24 Hz (Fig. 2) and from 13 to 33 at 48 Hz. The ratio of R, to R,,, ranged between 2 and 49% (mean: 16%) at 24 Hz and between 1 and 23 at 48 Hz. In the seven non-asthmatic infants R, varied between 1.1 and 10.6 cm H,0. L-’ - s at 24 Hz (Fig. 2) and between 0.6 and 7.3 at 48 Hz. R,,, ranged from 20 to

Nasal Impedance in infants 3

11 I

t (Y oe

60

5 0

40

3 0

2 0

10

0 asthmatic

infants

Fig. 2. Resistance of the nose (back row) and of the total respiratory system (cm H20 . L-’ . s) at 24 Hz including the nose (front row) in 30 infants.

39 cm H20 L-’ - s at 24 Hz (Fig. 2), and from 16 to 27 at 48 Hz. R, as a percentage of R,,, varied between 3 and 49% at 24 Hz and between 2 and 46% at 48 Hz.

Most frequently infants had an R, between 2 and 4 cm H,O - L-’ - s (Fig. 3). Non-asthmatic infants tended to have an R, in the same range as the asthmatic ones. A correlation was found between the nasal patency, as evaluated clinically, and R, (Fig. 4). Mean R, at 24 Hz for infants with an “open” nose was 2.5 k 2.0 cm H20 * L-’ * s-’ and was significantly lower (P < 0.05) than mean R, of the infants with a clinically “obstructed’ nose, which was 5.9 ? 4.0. However, there was an overlap between the two groups.

Fig. 5 shows R and X polynomials of two infants, one with a low and one with a high R,.

Mean difference at 24 Hz between R measured through the left nostril and through both nostrils was 2.7 -+ 4.1 cm H 2 0 - L-’ * s; between the right nostril and both nostrils this difference was 6.1 ? 7.4 cm H20 L-’ - s. In 17/30 asthmatic and 4/7 non-asthmatic infants the right nostril was dominant. Average frequency dependence (Fd) of R, of the asthmatic infants, expressed as the mean slope between 24 and 48 Hz was -0.09 -+ 0.08 cm H20 s2 * L (range: -0.29 to O.Ol) , whereas for R,,, Fd was -0.33 -+ 0.20 (range: -0.70 to -0.03) (P < 0.001). In the non-asthmatic infants, Fd of R, ranged between -0.20 and 0.00 cm H20 s2 * L and for R,,, between -0.53 and -0.02 (P < 0.05).

Average X, of the asthmatic infants was -0.9 k 3.7 cm H20 L-’ * s at 24 Hz and -0.1 * 1.9 at 48 Hz,

whereas mean X,,, was -16.0 i 8.2 at 24 Hz and 3.0 f. 1.9 at 48 Hz. Similar values were found in the non-asthmatic infants. In 19 asthmatic infants X, at 24 Hz was negative, in 6 it was positive, and in 5 zero (Fig. 6). In the group of non-asthmatics, six had a nega- tive X,, whereas 1 had a positive.

Finally, in seven infants the difference between two sequential R, determinations at a 45 min interval ranged from - 1.7 to 3.8 cm H20 * L-’ - s at 24 Hz and from -3:6 to 1 .O at 48 Hz. Changes in R, were not related to changes in R,,, (P > 0.05) (Fig. 7).

DISCUSSION

The present study suggests that, using a simple elec- trical model, nasal impedance can be determined by applying PRN oscillations, between 8-52 Hz, to the respiratory system through both nostrils and each nostril separately.

Measurement of respiratory system resistance in in- fants can be made with various methods.’Z2 Since infants are virtually obligatory nose breathers, all measurements are performed through the nose and values obtained include R,. Thus, results must be interpreted with caution, because an increased resistance can result from nasal obstruction. Moreover, it is not possible to compare values obtained in infants with similar measurements in older (mouth-breathing) subjects. With the PRN oscilla- tion technique adapted for infants, lo oscillations are

4 Desager et al.

15

4

8 a 5 E 3 C

0 0-2 2-4 4-6 6-8 8-10 * l o

Fig. 3. Number of infants (in 'YO of total) with increasing nasal resistance (0+10 cm H,O L-' s-I) at 24 Hz, represented as percentage of total respiratory system resistance (solid bars, asthmatic infants; shaded bars, non-asthmatic infants).

6 s t I

I I 0

0 norha1 redu$d Nasal patency

3

Fig. 4. Correlation between clinically determined normal and reduced nasal patency and measured nasal resistance (Ordi- nate, cm H20. L-' . s - I ) .

applied through a face mask covering the nose and part of the cheeks. This set-up is entirely different from the one in adults, where oscillations are applied at the mouth.8 It is assumed that the cheeks do not play a major role in infants. The measured R,,, also includes R,. In order to evaluate the effect of the nose, a method was needed to determine R, at the 13 forced oscillation frequencies. Since a large variability in R, has been reported, both between individual babies and between repeated mea-

surements in the same infant,',4 a relatively simple, non-invasive technique was required that could be per- formed routinely.

Although few investigators have reported measure- ments of R, in infancy, several methods have been applied in this age group. Using the esophageal balloon technique, Polgar' derived R, by subtracting pulmonary resistance measured through an oral airway from the values measured through the nose. R, has also been approximated by performing measurements through the left and right nostril separately and through both nostrils as in our study, but using the esophageal balloon technique4 or whole body plethysmography.6 Nasal oc- clusion was assumed not to alter the alternative nostril. Anterior' and posterior5 rhinomanometry have been adapted for use in infants, but technical difficulties were considerable, especially with the latter technique. Table 1 shows an overview of the absolute and relative values of R,, obtained with the above techniques. In our study R, at 24 Hz varied between 1 and 17 cm H20 . L-' * s and amounted to 2 and 4870 of R,,,. These values are smaller than those reported by others.'-' However, direct com- parison of the reported values is precluded for several reasons. First, the populations investigated are different. In the present study, 30 asthmatic and 7 non-asthmatic infants were studied, whereas others reported data for healthy infants. Since a major aim of our study was to evaluate the presented technique in a wide range of R,, measurements were performed in infants with a varying degree of nasal patency. Second, one must keep in mind that the PRN oscillation technique measures the resis-

Nasal Impedance in Infants 5

60

40

0 0

2o t P

10

0 1 L I . I .....

16 20 24 28 0 2 ¶6 4 0 44 48 62

rr .Ul lcy oh)

9 2o t 0

10

0 24 28 02 00 40 44 48 62

RycHnor ow Fig. 5. Typical resistance (cm H20 L-' . s) and reactance polynomials for an infant with a low nasal resistance (left) and an infant with a high nasal resistance (right). Respective lines are: total respiratory system impedance including the nose (-); total respiratory system imped- ances measured through the lefl nostril (--); and through the right nostril (- . -); nasal impedance ( . . . . ).

-50 I normal ll asthmatic

Infants

Fig. 6. Reactance (cm H,O . L-' . s) of the nose (back row) and of the total respiratory system (front row) at 24 Hz for 7 non-asthmatics and 30 asthmatic infants. Note that nasal reactance is low.

tance of the total respiratory system, whereas body plethysmography only assesses airway resistance (Raw) and the esophageal balloon technique evaluates airways plus lung tissue resistance. Therefore, R, as a percentage of Rtot should be smaller when measured by forced oscillometry. Absolute values of R, are also smaller than those reported by others. This can probably be explained

by the differences in technique; in our study resistance values were obtained at higher frequencies.

We have found it difficult to convince parents of normal infants to submit them to lung function testing with sedation. Therefore, it was not possible to collect data on a group of age-matched non-asthmatic infants. Values of only seven non-asthmatic infants are presented

6 Desager et al.

R24 (cm H2O.L-1.6) 40 I I

30 L

- . 1 2 3 4 5 6 7 Infants

Fig. 7. Changes in resistances (ordinate, cm H,O . L-' . s) at 24 Hz, of the nose ( . . . . .), and the total respiratory system including the nose (-) and excluding the nose (--) after a 45 min interval in seven infants.

and no statistics were performed on the comparison of asthmatics versus non-asthmatics. However, as expected, R, tended to be in the same range for both groups.

The comparison of measurements performed through one nostril and both nostrils provides a first rough estimate of the magnitude of R,. As shown in the left panel of Fig. 4, R values obtained from measurements through one nostril are nearly identical to those from measurements performed through both nostrils. This is explained by a usually small R,, relative to R,,,. In case of obstruction of one nasal passage, the R values, obtained from measurements through the obstructed nostril, increase (Fig. 4, right panel).

Since respiratory impedance values of infants show negative Fd, it was important to investigate whether the nose was responsible for this phenomenon. As our results show, there is no marked Fd of R,, in contrast with the marked Fd of R,,,. Therefore, measuring through the

nose only results in a parallel shift of the R,, polynomial to higher R values. Using the forced random noise oscillation technique the resistance of the nasal passages has been evaluated in normal adults and in subjects who were candidates for surgical correction of nasal obstruction." In the normals R, showed only minimal Fd, despite the large swings in the resistance of only one nostril, whereas the surgical candidates had a high degree of positive Fd of R,. In our study the infants with a higher R, showed a slight negative Fd. The contradictory results could be due to the differences in the anatomical struc- tures of the nose of adults and infants.

Validation of the assumptions made in the simple electrical model used in this study is not easy. Methods for evaluating R, in adults and children are technically difficult to use in infant^^.^ and the results obtained are poorly interchangeable with our technique. Calculation of R, by subtracting resistance measured with the PRN oscillometry through the mouth from resistance measured through the nose is another p~ss ib i l i ty .~ However, the size of any errors due to differences between shunt impedance during nasal and mouth breathing still needs to be evaluated. Moreover, in our study it was not possible to establish a free oral passage in sedated infants. Further measurements are needed to validate the current technique .

Prior to pulmonary function measurements nasal per- meability was evaluated by clinical inspection of the nose. This information was related with R,, as approxi- mated by our model and the latter was significantly higher in infants with an "obstructed' nose. Therefore, in infants without a history of upper airway obstruction or clinical evidence of nasal obstruction, R, can be expected to be low and to have only a minimal influence on R,,,.

Finally, changes in R, during a time interval of 45 min did not relate to changes in R,,,. This could be explained by the parallel arrangement of the two nasal air passages.

TABLE 1-Nasal Resistance in Infants

Population Methoda (cm H20. L-I.s) (9%) (%I Reference

Full-term neonates Esophageal 5.6- 19.9 42 11-41 Polgar and Kong3

Rnabsolute RnfRaw RJRp

(Two Caucasians + pressure (20-70) Three Negroes)

(Ten Negroes) pressure

Thirty Caucasians RMM 3.1-23.9 49 Stocks and Godfrey5 Thirteen Negroes 4.5-12.7 31

Thirty-three Caucasians RMM Volkheimer6. '

Full-term neonates Esophageal 4.2- 17.7 18-52 Lacourt and Polgar4

Infants Posterior

Infants Anterior 3.3-14.9 24 Lindemann and

Plethy smography 0.6-23.8 27 Infants PRN oscillation 0.8-16.6 2-49b Present study

Thirtv-two Caucasians ~ ~ ~~~~

aRMM, rhinomanometry; PRN, pseudo-random noise. RnIRtot.

Nasal Impedance in Infants 7

2. Milner AD, Saunders RA, Hopkin 1E. Relationship of intraoe- sophageal pressure to mouth pressure during measurements of thoracic gas volume in the newborn. Biol Neonate. 1978; 33:314-319.

3. Polgar G, Kong GP. The nasal resistance of newborn infants. J Pediatr. 1965; 6737-567.

4. Lacourt G, Polgar G. Interaction between nasal and pulmonary resistance in newborn infants. J Appl Physiol. 1971; 30:87&873.

5. Stocks J, Godfrey S. Nasal resistance during infancy. Respir Physiol. 1978; 34:233-246.

6. Lindemann H, Volkheimer C. Zur Messung des Nasenwider- standes bei Sauglingen. Atemw Lungenkrkh. 1981; 4:206208.

7. Lindemann H, Volkheimer C . Problems in measuring nasal resistance in infants and toddlers. Mod Probl Paediatr. 1982;

8. Landser FJ, Nagels J, Demedts M, Billiet L, Van de woestijne KP. A new method to determine frequency characteristics of the respiratory system. J Appl Physiol. 1976; 41:101-106.

9. Landser FJ, Polko AH, Visser BF. Oscillatory measurement of total respiratory impedance with extended spectrum up to 52 Hz. Arch Int Physiol Biochim 1983; 91:12.

10. Desager KN, Buhr W, Willemen M, LAndsCr FJ, Van Bever HP, De Backer W, Vermeire PA. Measurement of total respiratory impedance in infants by the forced oscillation technique. J Appl Physiol. 1991; in press.

11 . Fullton JM, Fisher ND, Drake AF, Bromberg PA. Frequency dependence of effective nasal resistance. Ann Otol Rhino1 Laryn- gol. 1984; 93:140-145.

12. Van Cauwenberge PB, Deleye L. Nasal cycle in children. Arch

2 1~47-52.

Obstruction of one passage should, therefore, only slightly increase R,. The changes in R, varied between 0.4 and 3.8 cm H20 - L-’ * s over a time interval of 45 min. To our knowledge there is no information on the natural variability of R, in infants. In 3-6 year old children the alternating cyclic activity, as found in adults, was absent. l 2 Resistances of the nostrils showed changes in the same sense, which were called “solidary move- ments” with a cycle of 57 min. It is not known whether this phenomenon also exists in infants.

In conclusion, nasal resistance and reactance can be fairly simply approximated from three consecutive mea- surements through both nostrils and through each nostril separately. Calculated nasal resistance at 24 Hz was found to vary between 1 and 16 cm H20 * L-l * s-’ and it showed no marked frequency dependence. Values of nasal resistance were related to nasal permeability as evaluated by clinical inspection. The nose was not responsible for the observed negative frequency depen- dence of the R,,,.

REFERENCES

1 . Doershuck CF. Matthews LW. Airwav resistance and lung volume in the newborn infant. Pediatr Ris. 1969; 3:128-134. Otolaryngol. 1984; 110:108-110.